INDEX

Explanations

code formatting indicators and syntax

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

edar

-0.19

Ãªs

-0.19

oÄŁunluk

-0.15

cona

-0.15

eda

-0.14

.BLL

-0.14

ji

-0.14

inger

-0.14

ly

-0.14

 Jacqu

-0.13

POSITIVE LOGITS

-transparent

0.16

`↵

0.16

askell

0.15

jun

0.15

ÙħÙĬ

0.15

bash

0.15

obs

0.14

umont

0.14

osh

0.14

Activations Density 0.004%

code formatting indicators and syntax

No Comments

No Known Activations